The Cohort Launch and Iterate Loop
- Difficulty
- Easy
- Time to result
- ~months to results
- Steps
- 7
- Confidence
- —
Dickie Bush's method for taking a product from idea to word-of-mouth engine, derived from how ship 30 for 30 was actually built. You start with whatever small audience you already have, float the idea and open a waitlist, then survey that list for the single thing people most want covered — which simultaneously validates demand and generates your content outline. You sell cheaply to a small first group, deliver it live, and then do the unscalable thing: one-on-one interviews with every participant asking what was best, what should be removed and what must stay. Each cohort's feedback rewrites the curriculum. After roughly ten cycles the problems that stopped people succeeding have been eliminated, the curriculum stabilises, and growth comes from participants telling their friends because the outcome is now reliable.
Origin
Extracted from Deep Dive with Ali Abdaal
How to run it
- 1
Start with the small audience you already have
A Twitter following, an email newsletter, any group you could plausibly offer the product to. You do not need scale to launch a first cohort.
- 2
Announce the idea and open a waitlist
Say you are thinking about launching the thing and invite people to join a waitlist if they would be interested. Interest at this stage is your first demand signal.
Pro tip Describe the offer as the specific way — for example a daily writing habit challenge.
- 3
Survey the waitlist for the one thing to cover
Ask each person what they most want you to cover, since you are still building it. The answers give you both the curriculum and the marketing language.
- 4
Sell a first cohort cheaply and deliver it live
Charge a small amount to a small group and run the thing in real time — calls, group chats, accountability. Live delivery surfaces the friction a polished product would hide.
Watch out Do not over-engineer version one; the point is to learn from delivering it.
- 5
Interview every single participant one on one
Sit in front of each person and ask why they joined, what made them want to quit, what the best part was, what should be removed and what must not be removed. Soak up as much information as the cohort can give you.
Pro tip Ask specifically about the moments people nearly dropped out — those are the constraints on completion.
- 6
Rewrite the curriculum against the feedback
Eliminate the problems that kept people from succeeding and double down on the parts people loved. Repeat across successive cohorts rather than trying to get it right once.
- 7
Stop iterating when the problems stop changing
After enough cycles the same issues stop appearing and the curriculum can be set in stone. At that point the product scales because every participant who transforms tells their friends.
Pro tip A product only scales if participants recommend it, so completion rate is the metric that matters.
In the wild
Dickie posted a 30-day accountability challenge on Twitter; fifty people paid $49 and he ran it with Zoom calls and group chats. Around 35 of 50 completed it, and he refunded essentially everyone out of nerves. He then held one-on-one calls with all fifty participants to understand what worked, what did not and what could be improved, using the interviews to build the real product.
→ The interviews became ship 30 for 30, which grew to 40K a month alongside his day job.
The team repeated the interview-and-rewrite loop for the first six or seven cohorts and kept iterating for about ten in total. By around cohort twenty the curriculum had not changed in roughly eighteen months, because the problems that blocked completion had been designed out. Participants who had wanted to write for forty years finally published, then told their writer friends.
→ A stable product with word-of-mouth distribution built in.
Common mistakes
Building the whole product before selling anything
Without a waitlist survey and a live first cohort you are guessing at the curriculum. The feedback that makes a product referable can only be gathered by delivering an imperfect version.
Skipping the one-on-one interviews
Aggregate surveys miss why people nearly quit. Sitting in front of individual participants is the only way to find the specific frictions that suppress completion and word of mouth.
Giving the product away to friends
Free participants almost never finish. Comped access removes the commitment that makes people act, so it produces neither results nor useful feedback.
From the transcript
“and this is actually how I advise people to launch a product”
“then I would send them a survey and say hey what's one thing that you really want me to cover doing this because I'm still…”
“and the only way to do that is through those one-on-one interviews where you really get to sit right in front of the person”
From the episode
How to Build a Business that Lets you Quit your Job - Dickie Bush
Dickie Bush